Depth-Synergized Mamba Meets Memory Experts for All-Day Image Reflection Separation

TL;DR

This paper introduces DMDNet, a novel deep learning framework that combines depth-aware guidance, state-space modeling, and memory modules to improve image reflection separation, especially in nighttime conditions.

Contribution

The paper presents DMDNet, integrating depth-guided saliency, state-space modeling, and cross-image memory modules for enhanced reflection separation, addressing nighttime challenges.

Findings

DMDNet outperforms existing methods in daytime and nighttime reflection separation.

The NightIRS dataset provides a new benchmark for nighttime reflection tasks.

Memory modules improve layer disentanglement by leveraging historical information.

Abstract

Image reflection separation aims to disentangle the transmission layer and the reflection layer from a blended image. Existing methods rely on limited information from a single image, tending to confuse the two layers when their contrasts are similar, a challenge more severe at night. To address this issue, we propose the Depth-Memory Decoupling Network (DMDNet). It employs the Depth-Aware Scanning (DAScan) to guide Mamba toward salient structures, promoting information flow along semantic coherence to construct stable states. Working in synergy with DAScan, the Depth-Synergized State-Space Model (DS-SSM) modulates the sensitivity of state activations by depth, suppressing the spread of ambiguous features that interfere with layer disentanglement. Furthermore, we introduce the Memory Expert Compensation Module (MECM), leveraging cross-image historical knowledge to guide experts in providing layer-specific compensation. To address the lack of datasets for nighttime reflection separation, we construct the Nighttime Image Reflection Separation (NightIRS) dataset. Extensive experiments demonstrate that DMDNet outperforms state-of-the-art methods in both daytime and nighttime.